The art of competing

Competition among companies and their products and services has not changed significantly in terms of the market dynamics (price, feature set, utility). What has changed is the pace at which feature parity among peers is converging. There used to be a real lasting advantage in developing specific features to differentiate from your competitors, but now those gaps are closed very rapidly. This morning I was reading Ulrich Homann latest blog post in which he explained this from a SaaS provider perspective and how many SaaS companies seem to be blind to their own competitive advantage (their moat). It seems weird but. it looks as if by focusing on your competitors product, you lose insight to what makes your product or service great.

That got me thinking: How does this translate to the company beyond the product, the people and how you use their capabilities well to stay in the game? To elaborate: How can you compete with your workforce in times where product advantage is always limited to the timespan in which the next idea becomes product, the latest insights appear and market volatility forces changing of the norms?

The value of people in IT

People were always the backstop for developing, using and operating IT systems, in some cases even the last resort. This was due to the nature that a human can reason beyond what could be programmed into the system reliably and cost-effective. Humans are uniquely capable when having to reason about incomplete, inconsistent, emerging and partial information in their day-to-day working context. They have true creativity and are strong in creating and working with novelty. This provides them the ability to act in ways that are hard to predict, especially without the full context and the learning experience the person performing the work has lived.

Human capital

On a side-note, there used to be a time when we called this “human capital”, to also express the value locked into the people working at a company in a manner that could not be expressed well on a spreadsheet. We seem to have lost this notion since the advent of AI/LLMs. It’ll be interesting to see if this will make a comeback due to the overemphasis on products and technology. Buy me a coffee if you want to hear the unfiltered story here.

More context does not equal broader utility

Anyway, when more and more specific skills and activities are translated into a program, this results in improved sophistication of action in a closed domain. It also increases the cost of the program and narrowed the utility provided to the specific domain or niche (it becomes non-transferrable), and ultimately you end up with specialty software.

The SaaS weakness

This is also why SaaS companies need to provide relatively generic functionality to be able to scale and generate enough margin to exist in their market. The only way to create transferrable software is to apply a generalization to a common problem domain. Don’t get me wrong, there may still be a lot of sophistication in the programs, but the knowledge required to “build this yourself” is also commonly available.

AI only scales for well defined systems

What we see, is that introducing a layer of AI in your system or way you operate, in some cases does bring productivity gains and reduces toil. This seems to work specifically well for performing work that is well defined (think: “closed domain, verifiable, enabled for automation”). Whenever we try to apply AI in more open-ended systems where distribution of input and outcomes may vary broadly, AI systems still have a hard time and you possibly have to invest a lot more in AI systems than you may anticipate from day one or what listening to the buzz-word-bingo may have suggested. From my perspective this is where I’d introduce AI augmentation instead of full task delegation, the strength of humans and the expertise they can bring is not to be underestimated.

Growing expertise

This brings us to an interesting topic, how do we grow new expertise and not have to rely on all our seniority having to keep the lights on. One of the imo crazy symptoms seems to come from the more senior developers now being at 100% utilization all the time. There is no slack, no room for knowledge transfer, just 100% focus on trying to figure out how to 100x their own productivity by learning to deal with agents.

On training and educating

This used to be very different. I remember when I spend Months together with juniors entering the field of development and spending time with them. We used to do a bit of tcp; train, coach, pair. Spending time, explaining how you see things, let them explore, guide their thinking and analysis, provide feedback on code, help them learn. However, the past two-to-three years have produced big challenges for early in career developers and people just getting started trying to land a job and progress. No one seems interested to spend the time with them to help them get on their feet. Everyone is distracted and chasing vanity metrics for productivity. I think this way to operate is chasing short term success and is losing the longevity of a company.

Sustained effort over time

When was the last time you thought about the next decade of your company and how to approach that? The markets seem interested in the next quarter maybe next 12 Months but nothing beyond that. I think for technology leaders that should be a warning, you’re under-investing in your future if all you think of is the next few quarters. Of course the startup ecosystem is brutal, financial runways run out in 18 Months or less, so in that space the dynamics and focus are more short term, but still a long term goals, true North Stars, help to find the right balance or harmony as Jeff Bezos calls this between going fast and well.

To go fast you must go well

The true productivity volume shows up as sustained effort over time. Agentic AI helps here with potential 24/7 work delivery. what matters more is “going well” that is to say: are you doing the right thing and are you doing the thing right!. This is where you need the right people, with the right skills, judgement, mindset and provide the ability to grow new juniors into well-rounded experts and senior technologists. The cost of doing that may seem huge, but not doing that is a guaranteed long-term failure and the key to unlock long-term competing power. Let’s build for generations to come.

As always, let me know what you think!